Life Tree
17 turning points — what he chose vs what he didn't
Each turning point pulled in three directions: the engineer path, the pure-art path, the comfort-zone path. This tree lists what he chose — and what he didn't. Every alternative comes from original-text footnotes or structural inference of the same generation. Source: Taiwan.md /people/cheyuwu/ self-portrait.
Born in Taipei into an AutoCAD family
Father Wu Yung-chin and mother Lin Mei-ying co-run Hung-Hong AutoCAD Technology Center. The household is a native technology environment, technical books constantly cycling through. What his parents gave him was not industrial-drafting technique — it was an unconditional safety net.
In elementary school, his mother brings home a Flash tutorial book from work and leaves it lying around
By sixth grade he is building the class website. His father picks up the relay and teaches him Visual Basic. The computer is not something he learns to use as he grows up — it is a native tool. The first brick of the teaching flywheel.
Most elementary schoolers of his generation, seeing a technical book at home, do not pick it up on their own. If he had stopped at the observer position of "what my parents do for work," the entire engineering competence line would be delayed by years.
Lin Mei-ying's original line: "Playing other people's games gets boring quickly. Playing games you designed yourself — that's the real thing." The fact that this line exists is reverse evidence that the path was available.
In middle school he extends the Visual Basic line from his father all the way to VB.NET, and writes a biological simulator
Virtual dots move, breed, forage and die on the screen. Seventeen years later these dots will swim onto a white wall at the Venice Biennale as Soul Fish. The middle-school work and Venice are the same line.
Most VB-literate middle schoolers of that generation made small games or forums. A biological simulator was a niche direction — no practical goal.
The other recognized path for strong-coding Taiwanese middle schoolers. Time spent on competition grinding would not have been left for this not-yet-named field of "algorithmic art."
First entry into the Acer Digital Creativity Award, in his second year of middle school
As a high-school freshman in the general division he beats the professional division and wins the jury's super grand prize. As a senior he teams up cross-school cross-division with Chen Kai-chung and wins again. Stan Shih hands him the trophy on stage; ten years later writes his NYU recommendation letter; thirteen years later sits in the front row of his solo show; sixteen years later shares the stage in conversation. A foreshadowing line that runs sixteen years.
Most prize winners enter once and pivot to other competitions. Had he stopped after the first time, no long-term thread to Stan Shih, no NYU recommendation letter line.
Most of his three years at Taipei Jianguo High School are spent walking alone. "When I got to Jianguo, I suddenly felt very defeated."
Specializes in chromatic harmonica. 2012 first prize in quartet at the National Student Music Competition with Czardas. The habit of "non-mainstream club is also belonging" later shows up in his choice of Art Blocks over mainstream galleries, of Hahow over an academic-tenure track.
Mainstream Jianguo clubs are debate, MUN, student government. Strong STEM peers usually went to debate to build college-application credentials.
The Taiwanese gifted-student path: spend all hours on coursework and competitions. Had he gone this route, no twenty-year harmonica-then-piano foreshadow line.
Admitted to NYCU Department of Electrical Engineering. The most orthodox STEM path in Taiwan
He knew from day one he didn't belong there — but he picked up the engineering foundation in signal processing, linear algebra, programming. The late nights in the dorm making a thousand lines breathe in code became the underlying vocabulary of his generative art.
Some of his generation, finding "I don't belong here," transferred to design school or self-studied. Had he taken this path, he wouldn't have the EE engineering training — the "fully programmatic generation" capability of the Art Blocks era would be discounted.
Classmate mainstream path: graduate school → TSMC / MediaTek. Stable, high-paying, predictable. Had he gone this route, the entire art line never happens.
During college takes freelance work under "Mo Yu Interactive Design," at 21 launches his first online course on Hahow
Hahow course later becomes 22,000+ students / three five-star courses / NYCU lecturer appointment. Teaching and creating are not a trade-off — they are the same flywheel.
Most college-age freelancers treat teaching as "something I'll do later." Had he delayed teaching by ten years, no NYU TA experience and no closed loop with the 2024 NYCU adjunct lecturer appointment.
The mainstream overseas path for Taiwanese STEM students. Had he studied full-time, no teaching brand built up at 21 — but maybe a smoother NYU outcome.
With Stan Shih's recommendation letter, enters NYU Tandon's Integrated Digital Media MS program
Simultaneously serves as TA for the Creative Coding course, sitting in the Brooklyn office every week. One student stuck around for two semesters; another took 750 minutes from HTML to a finished JavaScript project. The trait of "never give up on any student" crystallizes here.
The orthodox path for an EE undergrad. Career curve more stable, salary higher — but the "art + engineering intersection" training of Integrated Digital Media never happens.
At 21 he already had a freelance studio — entrepreneurship was viable. But without those two years of New York perspective (Manhattan Outernets, Brooklyn NYU, the New York art scene), the international-exhibition trajectory later wouldn't connect.
Strange Robots' code is lifted 90 percent verbatim (the comments not even removed) and listed for sale on Art Blocks. The generative-art community erupts
Art Blocks founder Snowfro arbitrates a 50/50 royalty split. The community renames the copies "hams," reframing them from "knockoffs" to "an apology and thanks to the original author." Top artists like Dmitri Cherniak send hams in support. The "Father of Hams" title is born this way.
The American legal path — find a lawyer, send cease & desist. Possibly winnable but enormous time cost, and community relationships burned. Many copied artists go this route and end up two years deep in litigation with no cultural capital to show.
The other conservative path: pretend not to see, avoid conflict. Had he chosen this, no Snowfro intervention, no hams culture, no "Father of Hams" title.
Leaves Outernets after 18 months — Manhattan startup
Conclusion is simple: at someone else's company, it's very hard to make your own vision actually happen. First full-time job, also the last.
PM path can ladder up to Director / VP. Manhattan salary plus stock options is a stable wealth path. Had he stayed, he'd become "a Taiwanese engineer doing ad-tech in New York" — not a Taiwanese artist on Art Blocks.
Google / Meta positions were open to him. Had he gone the "company hop" route accumulating credentials, the artist identity line would have been pinned down by the engineer identity.
Global Views magazine cover: "Under 30, Turned His Life Around with NFTs — Taiwan's Digital Designer with a Net Worth Over 100 Million NTD"
26 years old, 100M+ NTD, on the cover of a mainstream business magazine. The first wave's representative figure of Taiwan's NFT scene. Set up the dramatic contrast for the FTX zero-out that followed.
Some NFT high-net-worth peers chose to stay low-key, avoiding the label. Had he declined, the post-FTX "100M to zero" contrast narrative wouldn't have had as much material.
FTX Overnight Zero
Not a 50% loss, not a correction — overnight, zero. "When you've already satisfied everything you wanted to do in life, and you no longer have to worry about money — what is your meaning then? Suddenly, very empty."
The post-FTX choice
Shift from "control yourself" to "design a system that does not require you to control yourself." Eight-vault cash flow, automated transfers, financial constitution. "I am, in fact, grateful that event happened."
Most blowup victims pick this: leverage one more time and try to claw it back. The "I refuse to accept this" feeling is natural — but it tethers you to the gambling table with no exit.
The conservative path: leave the crashed sector completely, return to pure commercial design. Had he chosen this, no MonoLab later, no Universal Formula, no Venice.
[speculative] Peers who went "100M to overnight zero" have cases of multi-year mental low. The family safety net (no "I told you so") is the key variable making this alternative not happen.
Venice Biennale opening reception. An Italian curator asks: "Where can I learn about Taiwan? Like, really learn?"
No answer in the moment. Buried it in his body for two years. Opens the layer of "when AI learns Taiwan from the internet, whose Taiwan does it learn?" This question later becomes the seed of Taiwan.md.
English resources (Taipei Times / major news outlets / Wikipedia) plus academic papers — hand the curator a reading list and call it done. Most people take this path: polite, presentable, zero follow-through.
[speculative] Acute-reaction founder mode: ship a prototype within two months. Had he gone this way, the question wouldn't have "stayed in his body for two years" to settle into the SSOT framework — could have become another short-lived project.
Two years after Venice, the question finally surfaces
An open Taiwan knowledge base maintained as Markdown, CC BY-SA 4.0. 3/18 reported by Liberty Times, INSIDE follows up, NMTH becomes data-curation partner. 900+ GitHub stars, visitors from 100+ countries.
Could have been a cheyuwu.com subdomain — he curates alone, no PRs accepted. Low traffic, no contributors, no fork mechanism. Semiont is never born.
Taiwan knowledge + AI training data, in theory walks the SaaS path. Had he chosen this, content gets a paywall, AI can't learn it — violating the original "knowledge sovereignty" intent.
A Facebook post: "I've been seriously living with an AI assistant for two weeks"
Normal posts get 30–50 likes. This one breaks 6,000, with over 2,000 reshares. Muse reads all his notes, remembers his creative context and behavior patterns. From the original 15 thematic axes, a 16th grows: digital mirror and self-archaeology. Later spins off into the open-source project Semiont.
Keep Muse in his own Obsidian as a productivity tool. Had he done this, no 6,000-like collective resonance triggered, no Semiont.
Personal AI as a category went hot in 2024–2025. Could go the $20/month subscription route. But the keyword "symbiosis" would get diluted by product-market-fit pressure into "yet another AI assistant app."
The piano was put away in sixth grade to prepare for the basic competency exam. The harmonica club, the NYCU piano-club presidency in between — but "use the piano to write songs, publish them, let strangers listen" had to wait twenty years
Blue Horizon, Stars, StarTrack, The Other Shore of Dreams, The Bull and the Sudden Rain, Summer Migration. At Huashan's Algorithmic Verses he sits at the piano while code blooms behind him. Twenty years of silence crossed by a single public release.
Most peers who set down their instrument simply set them down. Picking it back up requires active decision plus public commitment (releasing on Spotify). Without restart, the twenty-year through-line never gets validated by his body.
A middle path: play for himself, for friends, never upload. But the boundary of "let strangers listen" is never crossed — twenty years of silence still sits there.